Frequency selective surface based bandpass filters in the near -infrared region.

Abstract

Frequency selective surface (FSS) based bandpass filters in the near infra-red region between 1--2 mum have been investigated and the results are presented. The filters consist of aperture elements arranged in a periodic lattice. The resonant wavelength can be varied by changing the dimensions of the elements, their center-to-center spacing and thickness. The intent was to integrate them onto a photodetector for color discrimination for an adaptive optoelectronic eye. Simulations of the filters have been carried out using two simulation tools---EMPiCASSO and FSDA_PRISM code. The latter permits the simulation of thick structures and is more suitable for this purpose. Aperture elements including circles, annular loops and squares were studied. Filters can also be sandwiched between multiple dielectric layers for improved spectral response. A key aspect of the thesis is that the metal acts like a plasma in the near-infrared region and therefore a new model was required to predict the spectral behavior. Multiple filters stacked on top of each other can provide bandpass characteristics with faster roll-off and a flatter reflection minimum. A design based on dual FSS filters with square apertures arranged in a square lattice and separated by magnesium fluoride layers is presented. This design was based on the new model incorporated into FSDA_PRISM code and measurements were conducted to validate the design. Fabrication of the filters in the near-infrared region on high index substrates like silicon/GaAs is demonstrated. The dimensions of the elements are in the sub-micron region and are patterned using a RAITH-150 electron beam lithography machine. Very high resolution features <50 nm are also demonstrated. The filters are characterized by measuring the spectral reflectance. The measurements are done for TE and TM excitations at 7&deg; off-normal incidence. We demonstrated good agreement with the simulation results.